NoVo: Norm Voting off Hallucinations with Attention Heads in Large Language Models

Open in new window